git-guardrails-claude-code
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates the installation of a PreToolUse hook for the Claude Code agent. This hook executes a local shell script (
block-dangerous-git.sh) to evaluate and potentially block shell commands before execution. - [PRIVILEGE_ESCALATION]: The setup process includes modifying file permissions with
chmod +xto allow the security script to run as a tool hook. - [INDIRECT_PROMPT_INJECTION]: The skill implements a processing layer for agent tool inputs to enforce safety policies.
- Ingestion points: The
scripts/block-dangerous-git.shscript reads JSON data fromstdinrepresenting the agent's proposed command. - Boundary markers: None identified; the script performs raw string matching on the extracted command.
- Capability inventory: The skill configures execution hooks within the agent's configuration files (
.claude/settings.jsonor~/.claude/settings.json). - Sanitization: The script utilizes
jqfor structured data parsing andgrepfor pattern-based filtering to identify and intercept restricted command strings.
Audit Metadata